·Research Articles·

Eat, Drink and Sell: Evidence from China

Jiuli Huang, Yanling Wang

Abstract: Operating in an environment with epidemic corruption, firms often engage in bribery to “get things done”. It is especially the case in China, where domestic market is segmented by provincial borders arising from protections by local governments. This paper studies firms' corruptive behavior of treating government offi cials lavishly to help them overcome local protections in their sales efforts. In so doing, we utilize a two-step procedure to correct firms' self-selection of sales entry. We find that corruption helps firms overcome local protection to enter other provincial markets, and once in these markets, corruption alleviates the negative effects of local protectionism on firms' sales volume.

 

Keywords: Eat-Drink, Corruption, Market Entry, Local Protectionism

JEL Classifi cation: D22 F1

1 Introduction

Firms' engagement in corruption acts, with government officials and among themselves, perhaps has lent the largest social support for governments worldwide to enact anti-corruption regulations. Common acts of corruption,such as bribery and fraud, are illegal virtually everywhere (Noonan, 1984), and more countries have adopted the Organisation for Economic Co-operation and Development (OECD) guidelines against corruption by firms (Kaufmann and Wei, 1999). Despite the substantial efforts, corruption transactions are as prevalent as ever in most countries and even on the rise in others (Kaufmann and Kraay, 2008; Lambsdorff, 2006). Scandals and reports involving self-dealing top management teams are common, so are reports of the bribery of bureaucrats and elected offi cials (Djankov et al., 2008). Once an unethical practice is seen as “the way things are done”, it can readily become an unwritten rule of competition despite the costs incurred (Brass et al., 1998; Oliver, 1997), which is particularly true in China, given its political and cultural background (to be discussed later in detail). Research on the engagement of firms in illegal corruption transactions is growing (Kwok and Tadesse, 2006), but contemporary investigations have focused on country and broad firm-level influences on corruption behavior (Davis and Ruhe, 2004; Habib and Zurawicki, 2002; Uhlenbruck et al., 2006). This paper offers an interesting case to study firms' corruption to advance their sales into other provincial markets in China, and to examine how corruption by firms helps overcome provincial barriers to enter other markets and alleviate the negative effects on their sales of protection favoring local products.

Different explanations have been put forward to explain why firms choose to corrupt people in position. Among those is the residual control theory of firm (Grossman and Hart, 1986; Hart and Moore, 1990), which suits relatively well in the Chinese context. The residual control theory argues that actual behavior of a firm depends on who owns the residual rights to control the firm's assets, that is, “the rights to determine the uses of assets under circumstances that are not covered by contractual terms” (Foss and Foss, 1999, p. 4). Politicians or government officials can wield their residual rights by imposing rules and regulations on firms (Johnson et al., 1998). Although a firm's exposure to public corruption varies depending on the pervasiveness of national corruption and the frequency with which a firm's activities bring it into contact with government officials (Lee et al., 2010), as long as government officials possess residual rights, firms will resort to corruption to extract favors. This has become very clear in China.

China, despite being the fastest growing developing country in the world during the last three decades, has witnessed serious corruption problems by firms (and many other players), which have been frequently reported in the media. Just as the residual theory suggests, firms' engagement in corruption in China reflects both their operating environment and their response to the external pressure in order to achieve private purposes. Chinese governments at all levels have the authority to allocate state resources and to impose rules and regulations on firms. But governments' execution of rules and regulations are not always transparent, which could vary from case to case depending on many factors involved. That creates large room for firms to corrupt government offi cials for their private gains. Corruption can function, to a certain degree, as a mechanism for resource reallocation in China, though it creates unfair competition and fosters more corruptive behaviors by those in power. More often than not, Chinese government offi cials expect bribes and grafts in exchange for practicing favoritism towards the individuals. Firms choose to corrupt offi cials to achieve“protection” and “green lights” (the “grease money”); or to prevent further grabbing by government offi cials (the “protection money”). The latter type is mainly for maintain a good relationship with government offi cials so that firms might not be prone to government interventions due to the many irregularities in implementing regulations.

In China, regarding firms' endeavor to advance sales in other provincial markets, they first have to overcome the hurdle arising from barriers across China's provincial administrative borders thanks to local protection. These provincial barriers have helped creating more opportunities for corruption in. Unlike in many other countries, intra-national entry barriers within China stem largely from local protectionism, that is, discrimination in favor of local firms by overt or covert local government policies or regulations. As reported in many articles, local governments in China have strong incentives to protect their firms at the expense of firms from other domestic regions (Bai et al., 2004; Bao et al., Forthcoming), and local protectionism takes on many forms. For example, local government agencies may try to increase the cost of firms from other domestic regions with stringent taxation inspections, unnecessary product standards or extra sanitation requirements. Local governments may even intervene directly in transactions involving firms from other regions. In particular, many local governments often revert to some “underground” or hidden discriminative measures so as not to contradict the endeavors by the central government to promote regional integration. Thanks to these overt and hidden protection measures, the whole Chinese market is in effect segmented into different provincial markets, which has a border effect as close or higher to those among European countries or those between the United States and Canada (Poncet, 2003). Facing discrimination from other regions, firms often resort to bribe government officials and those in power in destination markets to exchange for the “green light” to sell in their markets and to alleviate the degree of local protection on their sales volume. Corruption among firms and with government officials have become so entrenched in China, a country deeply obsessed in business dealings of building networks and social ties.

A challenging task, however, to examine the effects of corruption on firms' sales is to get accurate measurement on corruption at the firm level. Researchers have tried different ways to proximate (estimate) corruption (Reinikka and Svensson, 2006; Olken and Pande, 2012), including survey of perception (of corruption), bribery survey, estimates from direct observation, estimates by subtraction (of the amounts before and after corruption), and estimates from market inference (Olken and Pande, 2012). The former three methods are frequently used in firm-level studies, but each with their drawbacks. Perceptionbased corruption measures are most cost-effective in getting corruption data, but its accuracy is often subject to suspect, as questionnaire quality depends on respondents' education and experiences, and on their collaborations and viewpoints. Direct observations, such as the descriptions of judicial cases or media reports, could be more reliable and detailed, but not widely available for most firms, and thus are more suitable for case studies than for empirical investigations. As to bribery survey, the quality of the questionnaires is even more sensitive in China. Entrepreneurs are extremely cautious when it comes to surveys on issues about its relationship with governments so as to avoid any possible retaliation by government officials. Plus, corruption behaviors are illegal, so firms themselves tend to cover-up these behaviors rather than to confess them. Thus, firms may refuse to answer the survey questions or lie about their corruption practices, a practice not only unique in China. The complexities indicate that even if there are direct measures on corruptions, their reliability and credibility are open to debate.

In lieu of these deficiencies, we follow Cai et al. (2011) and use firm's“Entertainment and Travel Cost” (ETC), the itemized cost under category“Management Expenses”. It is true that ETC also includes normal travel and entertainment costs, but using ETC to proxy corruption is justifi ed, especially within the Chinese context. First, ETCs include a wide range of itemized costs including expenses of meals (eat and drink, where alcoholic drink can cost a few hundreds of US dollars a bottle), entertainment, tourism, cigarettes and gifts. The accounting regulations are unable to determine the eligibility of the costs of the aforementioned items to be normal business costs or corruption expenses, due to the asymmetric information between the regulators and firms. Take meal expenses to illustrate the point. It is frequently reported in the media and is a common practice in China that firms treat government offi cials lavishly by dining in luxurious restaurants and drinking expensive alcoholic beverages, followed by spas and entertainments. These eat-drink and entertainment activities usually lead to deep and large-scale corruptions, as deals such as power-for-money and official-and-businessman collusions are often concluded over the expensive meals/entertainments. Plus, ETCs can reimburse a wide range of itemized costs, such as kickbacks, commissions, red envelops, bribery of large amount, and are often accompanied with inflated or faked receipts (Cai et al., 2011). Fully knowing the widespread existence of corruption in ETC accounting practices, China's taxation authorities have regulations setting the maximum ratio of pretax deductible of ETC to sales. To that end, firms would have less incentive to tell lies regarding their ETC expenses in surveys, since they are normally below the maximum ratio. Also, firms' entertainment and travel costs include all other costs, not included otherwise. Thus, ETC includes a wide variety of costs, itemized or not.

This paper is also related to a few studies which have studied corruption in international trade. For instance, Dutt and Traca (2010) found that corruption in importing countries has both the extortion and evasion effects on bilateral trade, and the latter holds true in countries with high tariffs. De Jong and Bogmans (2011) examine the impact on trade of corruption in general and trade-related corruption in particular. They found that corruption in general hinders trade, while bribe paying to customs boosts imports, especially in countries with ineffi cient customs. Thede and Gustafson (2012) show that intermediaries can bypass the barriers originated from corruption and provide a special channel for exporting. Olney (2013) fi nds that corruption increases the probability of indirect exporting via intermediaries, but decreases the probabilities of domestic sell and exporting directly by firms. Our paper compliments the previous studies, but focuses instead on the effects of corruption on intra-national trade within China. We examine ETCs on firms' sales expansion into other provincial markets, in the presence of provincial barriers. We borrow the heterogeneous firm theory in international trade (Melitz, 2003) to model firms' sales entry decision and sales experience outside their home provincial markets. In the first stage, firms will decide whether to enter other provincial markets or not. Based on the entry decision, in the second stage, we will examine the effects of ETC in alleviating the negative effects of local protection on firms' sales volume.

We use data from a detailed survey on firms' sales destinations for 2400 firms operating in China in 2003 by the World Bank to inform our study. After controlling factors affecting firms' sales in other provinces, including productivity, scale, distance to and purchasing power of the destination markets, we fi nd that, corruption denoted by ETCs signifi cantly increases firms' chances to sell in other provinces, and helps ease the negative effects of local protectionism on their sales volume once in those markets.

The remainder of the paper is constructed as follows. Section 2 describes the data, Section 3 discusses the estimation strategy, Section 4 reports the results and Section 5 concludes.

2 The Data

The data we use come from the World Bank Survey of 2400 firms operating in China in 2003. The survey was part of the World Bank's project of the World Business Environment Survey. They randomly chose 2400 firms located in 18 cities across 13 provinces in three major regions: the west, the middle and the east. These firms operate in 10 manufacturing and 4 service industries, and include the whole spectrum of ownerships of foreign-owned (FDI), state-owned (SOE) and collectively or privately owned. Most information in the questionnaire was collected for the year of 2002, but some of the questions were tracked back to 1999 or 2000.

Information about firm's market entry is contained in Section E of the Questionnaire (“Relations with Clients”). In the survey, firms are asked whether they have sold products in other provinces (“Does your firm sell products to other provinces? ”). If yes, then firms are asked to identify the specifi c provinces (among the 31 provincial-level administrative regions in China) (“If yes, circle all the provinces to which your firm sells product”). Further, firms are asked to report the impact of local protectionism on their sales (“whether regional protectionism of a province adversely reduces the amount your firm sells to those provinces”) with three choices: “no bad effects”, “moderate bad effects”and “serious bad effects”. From the survey, we construct an indicator on the impact of local protectionism on firms' sales volume, with “0” for no bad effects and “1” for moderate and serious bad effects. If firms' ETC spending could alleviate the negative effects on sales, then corruption helps firms overcome local protectionism.

We normalize ETC expenses as a percentage to firms' sales, ETC/Sales, similarly used in Cai et al. (2011), to purge out size effects across firms. Potentially, a firm can have up to 30 entries for their out-of-home province sales, with potential firm-destination pairs of 72000 (=2400×30) in the data. We exclude some observations with obvious logical errors. For example, some firms mistakenly treated their home provinces as “other provinces”; others answered the questions on market entry and local protectionism even if they did not report to have out-of-home province sales. We exclude these observations from our empirical study.

Table 1 reports summary statistics for ETC/Sales among firms. It is clear that ETC/Sales is higher for firms that have entered more provincial markets (Mean of ETC/Sales for “Yes” is signifi cantly higher than that of “No” at the 5% level with p-value of 0.033), suggesting that ETC helps firms overcome barriers to advance their sales in other provincial markets. Firms with higher ETC/Sales also experienced less negative effects from local protection on their sales (Mean of ETC/Sales for “No” is signifi cantly higher than that of “Yes” at the 1% level with p-value of 0.0007). This suggests that high ETC/Sales will help alleviate the negative effects from local protection on their sales volume.

3 The Estimation Equations

Most surveyed firms have budgets for ETC, but only a portion of them have entered other provincial markets. Essentially, it is a two-stage self-selection process at work. In the first stage, based on the budget of ETC and a host of other factors, a firm decides whether to enter other provincial markets or not. This is the Market Entry Equation in the sense of extensive margin. In the second stage, conditional on its entry, a firm reports its experience with local protectionism based on the perceived impact of local protectionism on its sales volume, the Local Protection Equation or the Impact Equation. Econometrically, we use the modified version of Heckman (1979)'s two-stage selection model by Helpman et al. (2008). The first stage includes all firms and models their decision to enter other provincial markets, with the dependent variable being an entry dummy indicator(Entry).The second stage estimates the intensive margin of sales conditional on the entry decision, i.e., including only those firm-province pairs with positive sales. Different from Helpman et al. (2008) is that our second-stage estimation is designed to capture the impact of ETCs in alleviating the negative effects of local protectionism on firms' sale volume rather than sale volume itself. The two-stage estimation equations are:

where f and p indicate firm and province (i.e. the potential destination market) respectively. Equation (1) is the first-stage selection equation, and Equation (2) is to examine the effects of ETCs on local protectionism. Technically, the twostage selection model requires some exclusion variables that directly affect firms' decision to enter other provincial markets, but not on the impact of local protection on firms' sales. We follow Helpman et al. (2008) to choose gravity variables,i.e.,trade costs(Trade Cost)and destination market potential variables (Market).They include the following.Distancefp-distance between the city where firm f locates to the capital city of a provincial market p; Borderfp - whether firm f's home province is a neighbouring province with the potential market p or not;MarketSizep-the market size of province p,proxied by province p's GDP;PPowerp-GDP per capita in province p,and CoastDistancep-the distance from the capital city of market p to the nearest port facility.

In addition to industry fixed effects, we also include a vector X to control for firm-level covariates which affect firms' sales entry and volume. They are described briefl y below, with summary statistics provided in Table 2.

lnTFP and lnSales/labor:firm productivity to control its ability to overcome trade costs and expand sales into out-of-home provincial markets, following the literature of heterogeneous firm trade. TFP is firms' total factor productivity (in logarithm) calculated using the Levinsohn-Petrin method (Levinsohn and Petrin, 2003). This LP-method can solve the issue of input simultaneity, but shrinks the sample due to missing variables on inputs. So, we also construct labor productivity lnSales/labor,defined as sales per employee(in logarithm)as an alternative.

lnFirmSize:measured as the logarithm of employment.Size is one of the most important factors behind firm heterogeneity (Bernard and Jensen, 1995). Larger firms enjoy the push-pull effects of market expansion than smaller firms. Large firms enjoy the economy of scale that enables them to sell their products with low prices, leading to some competitive advantages in marketing. Large firms are also motivated or even forced to expand aggressively to fi nd markets for their products. Both will give them leverage over small ones in market expansion.

AD and Marketing: firms'efforts to promote their products. Firms rely on different ways to promote their products, including internet, newspapers, posters, magazines, TV and radio. The questionnaire asks whether firms use the above means to advertise their products, not on spending. We construct AD as the number of channels used. For instance, if a firm uses TV and radio to advertise its product,AD is 2.We also construct a related measure,Marketing,the ratio of sales personnel in total employees, averaged over years 2001 and 2002.

Bamem: a 0/1 indicator on whether a firm is a member of a business association, with 1 yes, and 0 no. Being a member of an industry organization often enables firms to get access to very valuable information on markets. In addition, industry association can coordinate and facilitate as a third party between firms and different levels of governments. We expect that being a member of an industry association helps firms enter other markets.

Age:years after birth.Age can have two competing effects.On the one hand, older firms are more experienced and have built up their intangible assets that help firms' sales expansion outside their home provinces. But, older firms tend to be more conservative.So whether Age has a significant effect on firms'choice on sales destinations needs empirical investigation.

Te c h: the Ratio of engineering and technical personnel in total employees, averaged over years 2001 and 2002.This is to capture whether the different technological content of goods have different ability to overcome the negative local protectionism in sales.

ExportShare:share of exports in total sales.We include firms'export shares to control for the competing forces for firms' sales. When firms target overseas markets, they will commit less effort to expand into domestic markets.

SOE and Foreign:binary indicators to capture firms'ownership structure as state-owned firms (SOEs) or foreign controlled firms (FDI) respectively. Being state-owned might entail firms some advantages to various political resources to bypass trade barriers; while these resources are not available to foreign firms and private/collectively owned firms. But, foreign firms can take advantages of investment-promoting policies for market entry.

4 The Empirical Results

4.1 The Baseline Results

Table 3 reports the baseline results. To facilitate comparison, it reports in parallel the regression results of the Market Entry Equation and the Impact Equation. We choose to estimate these two equations separately rather than use the builtin Stata program “heckprob”, because we are more interested in getting marginal effects for both equations. That is, we estimate the Market Entry Equation first, calculate the corresponding inverse Mills ratio (IMR), and add that as another explanatory variable in the Impact Equation. The productivity variables are lnTFP in column (1a) and (1b), and lnSales/labor in column (2a) and (ab), respectively. Estimation results for industrial fi xed effects are omitted for brevity. Below, we will describe the results.

As expected, the coefficient on ETC/Sales in both equations are significant at the 1% level, which indicate that higher ETC spending increases firms' probability to enter other provincial markets and decreases the negative effects of local protectionism on firms' sales volume. The results suggest two things. One, for firms in China, corruption expenditure represented by ETCs helps firms enter other provincial markets: firms treat government offi cials lavishly so that they can overcome local protections and enter these markets. Two, corruption is used by firms as the “grease money” to increase their market size: firms use corruption to get around local protection (carried out by local governments) and alleviate the negative effects of protection on their sales volume (the negative and signifi cant coeffi cient on ETC).

As to the magnitude, our results are meaningful. Take the regressions with lnTFP to illustrate the point. In the Market Entry Equation 1(a), holding all other variables at their means, 10% increase of ETC-to-sales ratio from the mean leads to the probability of entering one more provincial market by 0.0020 points. Given that the predicted mean probability of firm's market entry is 0.1922, the 0.0020 points increase in probability translates into 1.04 % increase in the probability. In the Impact Equation (1b), 10% increase of ETC-to-sales ratio from the mean leads to the probability of experiencing negative impacts decreases by 0.0012 points. Given that the mean predicted probability of experiencing negative effects is 0.1539, the 0.0012 points decreases translates into 0.76%decreases in the probability of experiencing negative effects. Similarly, one standard deviation increase in ETC-to-Sales ratio, a big increase, leads to 21.61% increase in the probability of market entry, and 11.28% decreases in the probability of experiencing negative effects on sales from local protection.

There are interesting findings regarding firm ownership dummies, Foreign and SOE.Compared with privately/collectively owned firms (the base firms), foreign firms or SOEs have higher probability of entering other provincial markets. Also, the negative effects of local protection on sales are signifi cantly different from zero for Foreign firms and for SOEs,indicating that switching a firm from being privately/collectively owned to either foreign owned or owned by Chinese governments signifi cantly decreases its probability of experiencing negative effects on its sale volumes. To understand the different results, we need to look at the ownership characteristics among the three firm types. Foreign firms in China not only enjoy investment-promoting favorable policies, but also are generally superior to collectively or privately owned firms. These advantages make it easier for them to overcome the provincial barriers and enter other markets, not captured in the observables. Once in the markets, the status of foreign firms makes them less vulnerable to local protection, presumably because local governments across China have always been trying to attract FDI, and thus wanted to give them good impressions about local business environments. For SOEs, their bonds with different levels of governments, even if these bonds do not give them special assistance in firms' endeavor to expand sales in other provincial markets, could serve as a signal and make them more immune to local protection than private/collective firms.

Firms' productivity significantly helps them overcome trade barriers. Using either lnTFP or lnSales/labor generates similar results, indicating that firms with missing data for TFP do not drive the regression results. The fi nding here is consistent with those in the literature regarding intra-national trade costs and international trade costs. Higher productivity helps firms overcome fi xed costs of selling in other markets due to the entry barriers. Similarly, higher productivity alleviates the negative impact of local protection on sales.

As to other firm-level controls, we also get meaningful results. For instance, large firms are more likely to expand sales in other provincial markets, and large firms are less likely to experience the negative effects of local protection on their sales volume.

The gravity-type variables in the first equation implies that firms tend to enter nearby provincial markets, especially adjacent ones, and enter markets with both larger size and higher purchase power. Furthermore, it seems more likely for firms to choose destination markets far away from the coast, where market competition may be fi ercer than in the hinterland.

The market entry equation and the impact equation are related. The null hypothesis of independent equations is rejected at the 5% level as indicated by the estimated results for inverse Mills ratios. This indicates the validity of the two-stage processes that firms first make market entry decision, and then decide how much to sell in those markets.

4.2 Sensitivity Analysis

In the sample, there are 1069 firms that have no sales activities outside their home provinces, accounting for 44.5 % among all surveyed firms. For those firms, their firm-province pairs are all zeros for all their 30 non-home provinces. To see how our results change, here, we exclude those firms from the sample. This shrinks the comparison base by nearly half. And the first-stage question now becomes why firms choose to enter some provincial markets, but not others. Table 4 reports the corresponding results with lnTFP and lnSales/labor respectively.

It is clear that all of the previous conclusions hold (albeit with different magnitudes),except for Foreign firm ownership dummy.The coefficients are in general larger, re-enforcing previous conclusions. In particular, increases in ETC spending will significantly raise firms' probability of entering other markets. Other factors, such as distance, border, market size and purchasing power all have larger marginal effects.For Foreign firms,in the market entry equation, it fails to be even marginally signifi cant. It suggests that being foreign owned does not give firms advantages to choose one provincial market over the other, although once inside the market, being foreign owned does alleviate the negative effects of local protection on their sales.

Overall,the results are robust regardless of the sample size.Replacing lnTFP with lnSales/labor or excluding these firms staying at their home provincial markets only do not change the main conclusions.

4.3 The Interaction Effects between Ownership and ETC

Our analysis above reveals some particular advantages possessed by foreign invested firms and SOEs over collectively or privately owned ones. Indeed, the institutional background of these firm types in China may facilitate them in market entry as well as alleviate the negative impact of local protectionism. This subsection explores the interaction effects between firm ownership and their ETC spending on their market entry decision and sales to capture how ETC works differently with various firm ownerships. To that end, we introduce the product terms of Foreign and SOE with ETC/Sales,and include these terms in the equations.

Given that both the Market Entry Equation and the sales volume equations are nonlinear, calculating the marginal effects for the interaction terms are not straightforward. We follow Norton et al. (2004) and use their method to correctly compute the marginal effects of an interaction term involving a continuous variable and a dummy in a non-linear regression. Table 5 reports the marginal effects with“Foreign × ETC/Sales”and“SOE × ETC/Sales”both in the full sample, and in the reduced sample of those firms entering at least one out-of-home provincial markets. For brevity, Table 5 only reports the results of the variables in interest,with the overall marginal effects for ETC/Sales,Foreign and SOE.

In both the full sample and the reduced sample, in the Market Entry Equation, there is a positive and significant coefficient on“Foreign × ETC/Sales”,but a significant and negative coefficient on“SOE × ETC/Sales”.The coefficients on the product terms alone suggest that increases in ETC in foreign owned firms leads to higher probability of these firms entering other provincial markets, while for state-owned firms, it is just the opposite. Eat-drink is more sensitive to foreign firms than to state-owned firms. But, since there is a separate significantly positive coefficient on SOE,and a negative coefficient on Foreign, we need to calculate the overall effects for ETC, Foreign and SOE to see the total effects.The overall marginal effects for Foreign is 0.041,much larger than the baseline results(0.023),indicating there is an added effect between Foreign and ETC for foreign-owned firms to enter other provincial markets.For stateowned firms,the overall effect for SOE is 0.014,similar as the baseline results, indicating that there is no added effect between ETC and being state-owned.Being state-owned might give firms some advantages in entering other provincial markets,however,for foreign-owned firms,they may have torely more on ETC to pave the way by building up relationship capital with clients and government offi cials.

For the Impact Equation in both samples,the coefficient on Foreign × ETC/Sales is not significant, though positive, and is 5% significant (and positive) for SOE × ETC/Sales. But for overall effects, the marginal effects for neither Foreign nor SOE changed statistically, implying that once firms have entered the markets, there are no additional signifi cant added effects between ownership and ETC.

5 Concluding Remarks

China is a large country with segmented provincial markets by local protections.In the presence of market entry barriers, firms resort to corrupt Chinese government offi cials to advance their sales in more markets. The corruption to ease local protectionist barriers, with eat-drink the obvious one, has been the way to go for years. Behind the eat-drink facade, corruption of more severe forms follows, which are often shown as legitimate costs under itemized costs in firms' entertainment and travel costs. Firms expect to get favorable treatments from government offi cials so that they can sell their products in other provincial markets. This paper explores the eat-drink-sell relationship using Chinese firmlevel survey data by the World Bank.

We borrow the heterogeneous firm theory in international trade to explore this relationship. We model firms' sales in other provinces as a two-stage process. In the first stage, firms choose whether to enter other provincial markets or not. In the second stage, we estimate the effects of corruption to alleviate the negative effects of local protection on firms' sales. Our results indicate that spending on eat-drink helps firms get around local protections to sell in other provincial markets. The more they spend, they more likely they enter other provincial markets, and the lower the probability of experiencing negative effects of local protection on their sales. Eat-drink works better for foreign-owned firms than for state-owned firms.

References

Bai, Chong-En, Du, Yingjuan, Tao, Zhigang and Tong, Sarah (2004). “Local Protectionism and Regional Specialization: Evidence from China's Industries” , Journal of International Economics,63(2),397-417.

Baldwin, John, and Wang, Yanling (2011). “Plant Size, Nationality and Ownership Change”,International Journal of the Economics of Business,18(3), 351-380.

Bao, Qun, Huang, Jiuli and Wang, Yanling (Forthcoming). “Productivity and Firms'Sales Destination: Chinese Characteristics”, Review of International Economics.

Bernard, Andrew B., Jensen, J. Bradford (1995). “Exporters, Jobs, and Wages in U.S.Manufacturing:1976-1987”,Brookings Papers on Economic Activity:Microeconomics,67-112.

Brass, Daniel. J., Butterfield, Kenneth D. and Skaggs, Bruce C. (1998).“Relationships and Unethical Behavior: A Social Network Perspective”, Academy of Management Review,23(1),14-32.

Cai, Hongbin, Fang, Hanming and Xu, Lixin Colin (2011). “Eat, Drink, Firms and Government: An Investigation of Corruption from the Entertainment and Travel Costs of Chinese Firms”,Journal of Law&Economics,54(1),55-78.

Chen, Yunling, Liu, Ming and Su, Jun (2013). “Greasing the Wheels of Bank Lending:Evidence from Private Firms in China”,Journal of Banking&Finance,37(7),2533-2545.

Coughlin, Cletus C., Novy, Dennis (2013). “Is the International Border Effect Larger than the Domestic Border Effect? Evidence from US Trade”, CESifo Economic Studies,59(2),249-276.

Davis, James H., Ruhe, John A. (2004). “Perceptions of Country Corruption:Antecedents and Outcomes”,Journal of Business Ethics,43(4),275-288.

De Jong, Eelke, Bogmans, Christian (2011). “Does Corruption Discourage International Trade? ”,European Journal of Political Economy,27(2),385-398.

Djankov, Simeon, LaPorta, Rafael, Lopez-di-Silanes, Florencio and Shleifer, Andrei(2008).“The Law and Economics of Self-Dealing”,Journal of Financial Economics,88(3),430-465.

Dutt, Pushan, and Traca, Daniel (2010). “Corruption and Bilateral Trade Flows:Extortion or Evasion? ”,Review of Economics and Statistics,92(4),843-860.

Fisman, Raymond, Svensson, Jakob (2007). “Are Corruption and Taxation Really Harmful to Growth? Firm Level Evidence”, Journal of Development Economics,83(1),63-75.

Foss, Kirsten, Foss, Nicolai J. (1999). “Understanding Ownership: Residual Rights of Control and Appropriable Control Rights”, DRUID Working Paper No. 99-4.

Grossman, Sanford J., Hart, Oliver D. (1986). “The Costs and Benefits of Ownership:A Theory of Lateral and Vertical Integration”,Journal of Political Economy,94(4),691-719.

Habib, Mohsin, Zurawicki, Leon (2002). “Corruption and Foreign Direct Investment”,Journal of International Business Studies,33(2),291-307.

Hart, Oliver, Moore, John (1990). “Property Rights and the Nature of the Firm”,Journal of Political Economics,98(6),1119-1158.

Heckman, James J. (1979). “Sample Selection Bias as a Specifi cation Error”, Econometrica,47(1),153-161.

Helpman, Elhanan, Melitz, Marc and Rubinstein, Yona (2008). “Estimating Trade Flows: Trading Partners and Trading Volumes”, Quarterly Journal of Economics,123(2),441-487.

Hillberry, Russell, Hummels, David (2003). “Intra-national Home Bias: Some Explanations”,Review of Economics and Statistics,85(4),1089-1092.

Hillberry, Russell, Hummels, David (2008). “Trade Responses to Geographic Frictions:A Decomposition Using Micro-Data”, European Economic Review, 52(3), 527-550.

Huang, Jiuli, Li, Kunwang (2013). “Eat-Drink, Corruption and Firm's Purchase Order”,Economic Research Journal[In Chineses],48(6),71-84.

Jensen, Nathan M, Li, Quan and Rahman, Aminur (2010). “Understanding Corruption and Firm Responses in Cross-National Firm-Level Surveys”,Journal of International Business Studies,41(9),1481-1504.

Johnson, Simon, Kaufmann, Daniel and Zoido-Lobaton, Pablo (1998).“Regulatory Discretion and the Unofficial Economy”, American Economic Review,88(2),387-392.

Kaufmann, Daniel, Kraay, Aart (2008). “Governance Indicators: Where Are We,Where Should We Be Going? ”,World Bank Research Observer,23(1),1-30.

Kauffman, Daniel, Pradhan, Sanjay and Ryterman, Randi(1999). “World Bank Finds New Ways to Diagnose Corruption Symptoms”,Transition,10,10-11.

Kaufmann, Daniel, Wei, Shangjin (1999). “Does ‘Grease Money' Speed up the Wheels of Commerce? ”, NBER Working Paper No. 7093.

Kwok, Chuck C. Y., Tadesse, Solomon (2006). “The MNC as an Agent of Change for Host-Country Institutions: FDI and Corruption”, Journal of International Business Studies,37,767-785.

Lambsdorff, Johann Graf (2006). “Causes and Consequences of Corruption:What do We Know from a Cross-Section of Countries? ”, in Susan RoseAckerman(ed.),International Handbook on the Economics of Corruption,New York: Edward Elgar, 8-34.

Lee, Seung-Hyun, Oh, Kyeungrae and Eden, Lorraine (2010).“Why do Firms Bribe? Insights from Residual Control Theory into Firms' Vulnerability and Exposure to Corruption”,Management International Review,50,775-796.

Levinsohn, James, Petrin, Amil (2003). “Estimating Production Functions Using Inputs to Control for Unobservables”,Review of Economic Studies,70(2), 317-342.

Martincus, Christian Volpe, Blyde, Juan (2013). “Shaky Roads and Trembling Exports: Assessing the Trade Effects of Domestic Infrastructure Using a Natural Experiment”,Journal of International Economics,90(1),148-161.

Melitz, Marc J. (2003). “The Impact of Trade on Intra-Industry Reallocations and Aggregate Industry Productivity”,Econometrica,71(6),1695-1725.

Nitsch, Volker, Wolf, Nikolaus (2013). “Tear down this Wall: on the Persistence of Borders in Trade”,Canadian Journal of Economics,46(1),154-179.

Noonan,John T.(1984),Bribes, New York:Macmillan.

Norton, Edward C., Wang, Hua and Ai, Chunrong (2004). “Computing Interaction Effects and Standard Errors in Logit and Probit Models”, Stata Journal,4(2),154-167.

Oliver, Christine (1997). “Sustainable Competitive Advantage: Combining Institutional and Resource-Based Views”,Strategic Management Journal,18(9), 697-713.

Olken, Benjamin A., Pande, Rohini (2012). “Corruption in Developing Countries”,Annual Review of Economics,4,479-509.

Olney, William W. (2013). “Impact of Corruption on Firm Level Export Decisions”, Working Papers 2013-04, Department of Economics, Williams College.

Poncet, Sandra (2003). “Measuring Chinese Domestic and International Integration”,China Economic Review,14(1),1-21.

Reinikka, Ritva, Svensson, Jakob (2006). “Using Micro-Surveys to Measure and Explain Corruption”,World Development,34(2),359-370.

Thede, Susanna, Gustafson, Nils-Åke (2012). “The Multifaceted Impact of Corruption on International Trade”,World Economy,35(5), 651-666.

Uhlenbruck, Klaus, Rodriguez, Peter, Doh, Jonathan and Eden, Lorraine (2006). “The Impact of Corruption on Entry Strategy: Evidence from Telecommunication Projects in Emerging Economies”, Organization Science, 17(3), 402-414.

Wagner, Joachim (2007). “Exports and Productivity: A Survey of the Evidence from Firm-level Data”,World Economy,30(1),60-82.

Wolf, Holger C. (2000).“Intranational Home Bias in Trade”, Review of Economics and Statistics,82(4),555-563.

Xu, Chenggang(2011). “The Fundamental Institutions of China's Reforms and Development”,Journal of Economic Literature,49(4),1076-1151.

Young, Alwyn (2000). “The Razor's Edge: Distortions and Incremental Reform in the People's Republic of China”, Quarterly Journal of Economics, 115(4), 1091-1135.